From: Russell King - ARM Linux Subject: Re: [PATCH 1/9] ARM: Do early I/O mapping if spinlock debugging is enabled Date: Wed, 28 Jan 2009 18:53:55 +0000 Message-ID: <20090128185355.GC23301@n2100.arm.linux.org.uk> References: <20090128181931.21007.73744.stgit@localhost> <20090128182305.21007.62781.stgit@localhost> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from caramon.arm.linux.org.uk ([78.32.30.218]:55365 "EHLO caramon.arm.linux.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751489AbZA1SyJ (ORCPT ); Wed, 28 Jan 2009 13:54:09 -0500 Content-Disposition: inline In-Reply-To: <20090128182305.21007.62781.stgit@localhost> Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: Tony Lindgren Cc: linux-arm-kernel@lists.arm.linux.org.uk, Juha Yrjola , linux-omap@vger.kernel.org On Wed, Jan 28, 2009 at 10:23:05AM -0800, Tony Lindgren wrote: > From: Juha Yrjola > > At least on OMAP, sched_clock() requires the I/O maps to be initialized. > Spinlock debugging invokes sched_clock() very early. NAK. This doesn't and can't work on every platform out there. The only viable fix is that sched_clock() must NOT be used early. If the kernel is using sched_clock() early, that's a bug.